Attributions & Licences

Apollo's curriculum is built on top of freely licensed Japanese linguistic data produced by other people. We are grateful for it. This page credits those sources and links the licences that govern them.

Dictionary and linguistic data

JMdict

Japanese–English dictionary data from the JMdict project, property of the Electronic Dictionary Research and Development Group (EDRDG), Monash University. Used under the EDRDG Licence and CC BY-SA 4.0.

Changes we made: entries were parsed, filtered to a study-relevant subset, assigned to our own JLPT level distribution, reshaped into flashcard and quiz records, and paired with generated distractors and audio.

KANJIDIC2

Kanji character data — readings, meanings, stroke counts, and classifications — from KANJIDIC2, property of the EDRDG. Used under the EDRDG Licence and CC BY-SA 4.0.

Changes we made: converted to our own record format, filtered, level-assigned, and enriched with primary-reading selections and teaching notes we authored.

KRADFILE / RADKFILE

Kanji-to-radical and radical-to-kanji decomposition data, property of the EDRDG, originally compiled by Jim Breen. Used under the EDRDG Licence and CC BY-SA 4.0.

Changes we made: reindexed to support component-based search and study grouping.

KanjiVG

Stroke-order and character-shape vector data by Ulrich Apel and the KanjiVG project. Used under CC BY-SA 3.0.

Changes we made: paths were extracted and re-rendered to drive Apollo's handwriting practice and stroke-order animation.

JmdictFurigana

Furigana reading alignments from the JmdictFurigana project, derived from JMdict and KANJIDIC. Used under CC BY-SA 4.0.

Changes we made: merged into our vocabulary records to position furigana over the correct characters.

Tatoeba

Some example sentences reaching Apollo through JMdict originate from the Tatoeba Project, licensed CC BY 2.0 FR.

What is our own

For clarity about what the licences above do and do not cover: Apollo's JLPT level distribution for kanji and vocabulary is our own work, edited and maintained by us rather than derived from an existing JLPT list. Our lesson text, grammar explanations, exercises, deck structure, teaching notes, and the application itself are likewise original or licensed separately.

Audio

Spoken Japanese audio in Apollo is synthesised with Google Cloud Text-to-Speech, using the WaveNet voice ja-JP-Wavenet-A, under the Google Cloud Platform Terms of Service.

Those terms (§5.1) provide that the customer retains intellectual property rights in their own data and applications, and Google claims no rights in the generated output. No attribution is required for the audio — this credit is a courtesy, not a licence condition.

Third-party services

Reading stories are generated using Google's Gemini API. Optional progress sync is provided through WaniKani by Tofugu LLC, used only when you connect your own account.

Software

Apollo is built with Flutter and Dart (BSD-3-Clause), Go (BSD-3-Clause), PostgreSQL (PostgreSQL Licence), and a number of open-source libraries under permissive licences including MIT, BSD, and Apache-2.0. A complete dependency listing is available on request from support@apolloapp.site.

Corrections

If you believe something here is credited incorrectly, or that content in Apollo is used without proper licence, please tell us at legal@apolloapp.site and we will investigate promptly.